Forensic Analysis for Source Camera Identification from EXIF Metadata.

Area of Science:

• Digital Forensics
• Multimedia Forensics
• Computational Photography

Background:

• Source camera identification is crucial for digital evidence verification and copyright protection.
• Photo-Response Non-Uniformity (PRNU) is a reliable method for device-specific fingerprinting.
• Modern computational photography causes anomalous behaviors in PRNU analysis, challenging existing methods.

Purpose of the Study:

• To develop a novel framework for source camera identification that overcomes limitations of current PRNU-based approaches.
• To investigate and address anomalous device behaviors in computational photography impacting forensic analysis.
• To leverage Exchangeable Image File Format (EXIF) metadata for enhanced camera identification.

Main Methods:

• A forensic analysis framework utilizing Exchangeable Image File Format (EXIF) metadata was developed.
• EXIF metadata was represented as type-aware word embeddings to preserve contextual information.
• The framework was designed for visual interpretation and identification of anomalous device behaviors.

Main Results:

• The proposed method achieves state-of-the-art performance in source camera identification.
• The framework provides valuable insights into anomalous behaviors in modern smartphone cameras.
• Experiments on public datasets validate the effectiveness and robustness of the approach.

Conclusions:

• The EXIF metadata-based framework offers a simple yet effective solution for source camera identification.
• This approach enhances the reliability of multimedia forensics in the era of advanced computational photography.
• The method provides a pathway for understanding and mitigating anomalies in device-specific fingerprints.